New "sting" of weak open-access journals.
 
This afternoon John Bohannon published an article in Science exposing lamentably low quality at a large number of OA journals. He and Science call it a "sting" and it's easy to see why. Unfortunately it may be hard to disentangle what the article does and doesn't show. My take:

* It shows that many OA journals are weak. 

* It doesn't show that all or most OA journals are weak. Bohannon didn't study all or most OA journals.

* It doesn't show that overall or on average, OA journals are weaker than non-OA journals. Bohannon didn't study non-OA journals.

* It shows that some OA journals are strong. In particular, it vindicates peer review at Hindawi and PLoS ONE. I'm particularly pleased by this result because Hindawi has been unfairly characterized as "predatory", and because PLoS ONE has been unfairly characterized as using lax peer review. PLoS ONE evaluates articles for methodological soundness and not significance. That's less than conventional peer review takes on, but it doesn't mean, and never meant, that PLoS ONE couldn't be as rigorous as any other journal in evaluating methodological soundness.

* Putting some of these together, the article shows that some OA journals are weak and some are strong. 

* It doesn't show that the problem it exposes is limited to OA journals, or intrinsic to OA journals. It exposes a problem with low-quality or dishonest journals, not with OA journals as such.

* It doesn't show that low quality is non-existent or rare on the subscription side of the line. It merely singles out low quality on the OA side of the line. 

* It doesn't show that good OA journals cannot be as good as the best non-OA journals.

* It doesn't show anything about green OA, or OA delivered by repositories. Hence, it doesn't expose a problem intrinsic to OA itself, which is broader than gold OA or OA delivered by journals. It doesn't even mention green OA. Nor does it mention the fact that researchers can publish in the very best journals in their field and deposit a copy of their peer-reviewed manuscript in an OA repository.

But apart from what the article does and doesn't show, it will have consequences for the perception of OA journals and the perception of OA itself. I'm afraid it will have these consequences apart from what the article does and doesn't actually show.

One effect is good, and I'll mention it first. Weak and dishonest OA journals give OA a bad name. I want to expose them. I want to warn authors and readers against them. I want to drive them from the field. To some degree, Bohannon's paper will help in that cause.

Unfortunately it will probably overshoot. Many people incorrectly believe that all OA journals are weak and dishonest. Hence, many will put all OA journals under the cloud of suspicion. Many people incorrectly believe that all OA is gold OA, or that OA journals are the only way to deliver OA. Hence, many will put all OA, not just gold OA, and not just the weak subset of gold OA, under the cloud of suspicion.

Bohannon is not responsible for these widespread, pre-existing misunderstandings. But his conclusions combine badly with them, especially when he is not careful in drawing his conclusions or in characterizing OA. 

He refers to "an emerging Wild West in academic publishing" as if low-quality journals were something new. This is unjustified and invidious, especially since he chose to study only OA journals, which tend to be new. Because he deliberately omitted to study non-OA journals, he should carefully avoid the conclusion that the kind of low quality he has exposed is something new.

In two places he gives the false impression that all or most OA journals charge publication fees. But most charge no fees at all. It's possible that all or most of the OA journals in his sample charge publication fees. But he doesn't say so and doesn't seem to have checked. He makes it easy for readers to draw unwarranted conclusions about OA journals as a class and about the fee-based business model for OA journals.

http://www.sciencemag.org/content/342/6154/60.full

#oa #openaccess  
Shared publiclyView activity